*OFFICIAL* Alienware m15 Owner's Lounge

I did a dumb little test for anyone that cares.
FF15 1080p High benchmark :
1300 Core across - 6822pts, 76.929W (didn't check temp)
1515 Core across - 7651pts, 86.556W, 79C
Stock curve - 7659pts, 91.972W, 80C

Can't confirm the stability being 100% but I ran 1515 core across twice fine. Kinda seems like, other than making the GPU thermal throttle less, there really is no room for overclocking or undervolting since the GPU is already very efficient. On my RTX 2080 and the 1070 of my 15R3, you are able to undervolt + overclock which will get stock-like performance (or more) with a way lower power usage. With the 2070 Max-Q, it just seems like the wattage goes down a small amount and you take a huge performance hit.

Also, I can't confirm it until I do the PCH mod, but my PCH is hitting 91C and I am stuttering in RB6 Siege randomly sometimes. Wonder if it is related to the PCH, since there is nothing else that could be causing it.

 

Most likely Dell's engineers Jokemen fiddled with PCH Core Voltage for the platform controller hub. None knows. As well there is a difference in the binning. Some PCH chips will run hotter than others in same way as two processors with equal voltage don't necessarily reach same temp at same wattage.

 

Thermal paste doesn't really stick. It's just a barrier to prevent that LM goes out. We had some discussions if thermal paste and LM togther have disadvantages or reactions. I tested it over months with different alienware heatsinks. (test: destroys LM my heatsink). I couldn't test anything negative.

That's right. I suspect that 5mm is still too little.

I might take something else. I did a test ( full test) with different thermal pads and thermal adhesive tape and the tape performs really poor. Look in chart from the link. (thermal adhesive tape = Wärmeleitklebeband)

It can be. I have a lot of HWinfo logs from my R5 and and I always test under the same conditions. After a BIOS update, my watts under load have risen (~ 10w more) and the PCH temps have risen.

 

That's all I could find on Amazon. I guess there are different kind of tapes with different characteristics. With this one they claim 1.5W/K*m and I picked the thinner 0.15mm. I really prioritised strong adhesive properties. Of course it won't be as good as the paste, but I couldn't even find a large 3mm high heatsink to consider the glue route (actually there was one, with delivery from China in a month).

I'm pretty sure this will do. I used the laptop for 5 hours yesterday, avg PCH temp 60, max 70C (no benchmarking).

 

I tried the thermal adhesive with bonding 2 heatsinks and idk if I used too much glue (tried to use about the same amount as you did) but they aren't easily coming apart. Maybe if I used a very tiny amount it would both hold well + be remove-able but Idk if I want to risk that. Apparently Kapton tape holds extremely well so I am kinda considering maybe I can use that to keep the heatsink on the PCH lol

 

I'm going to do the repaste and PCH cooling mods today. Just wanted to capture the initial state 6 months after my Kryonaut repaste. Tests were done at 22C ambient.

CB20


The CPU was suffering from hugely uneven temps - 17C max difference:

This phenomenon was noticeable right after the original repaste, but it was more in the ballpark of 8-9C.

PCH temp is a lol during CB20 benching:


TimeSpy (431.60 driver)
Kryonaut continues to do a great job on the GPU, there is no reason to repaste this (but I will try Phobya for fun, rather than re-apply Kryonaut). PCH still pretty much lolled at 64C max.






Sandra Lite Memory Test
HyperX Impact delivered the goods, while the PCH continued to ridicule us with 48C max temp.




AS SSD
I also ran a quick test of my 970 EVO Plus 1TB SSD to check interaction with the PCH and whether the temps are OK.





Basically, in all those tests the PCH laughed very very hard at my pitiful testing efforts. Yet, perhaps I could have done better with additional PCH cooling - we will soon find out for sure!

I will post my post-repaste results next, and then follow up with results on the effects of PCH and RAM cooling improvements. Stay tuned!

 

Overall, I'm relatively happy with the results of Phobya repaste. The results are noticeably better, and the 8750H performs pretty close to max performance.

CB20 (+3%)
While Phobya didn't resolve uneven temps, the temps came down quite a bit. This allowed me to increase the power, and thus the higher result. The PCH worked at a bit higher temp, but that still amounted to max 59C.







TimeSpy (+0.4%)
The GPU performed identically as with Kryonaut, the increase in score was entirely due to the CPU component. Interestingly, max PCH temp went up to 80C.

 

Then I went back to pages 314-317 and applied @kahuna0k 's static voltage trick. It's completely ridiculous - temps down several C while the CB20 score went even higher.





tPCH was 56C max.

 

Finally, the big moment! I applied RAM (3mm gelid extreme) and PCH cooling (2x20x20x1.5mm copper shims stuck on top of the PCH + 3mm gelid extreme on top for case contact to see if it will have any effect on performance:

CB20
Exactly the same score, which makes sense because the CPU is very close to maxing out anyway, however the PCH indeed ran cooler (about -8C max):


TimeSpy
Same score.

Sandra Lite Memory Test
Exactly the same result, and strangely the same PCH temp.

Executive summary: extra cooling of the PCH has absolutely no impact on performance at 22C ambient.

 

What do you expect from a load which goes 1:30? play a game for min. 30min or more. A small benchmarks shows not really much. Time spy goes 2 minutes.
Same for your HWinfo logs. No time, no avg package power, no uv-values and so on.

 

Well, any significant effects would show up, TS runs longer than 2min, but what did we expect to happen if the PCH runs at 60-80C in the first place. Absolutely nothing, it's basically a fun waste of time if one is into modding.

(Package power in CPU tests was 65W during short TB and 63W during long TB. Nice straw man lol)

 

Why don't you show the whole HWinfo log? Do you have something to hide or what is the reason?

I logged my values on the same way like you and i've surprisingly a big improvement of the temperatures.




My new temps in TS and CB

 

Sorry, something's seriously wrong with you. I mean who cuts a big hole in the laptop to stick a bunch of radiators to his heatsink... Dude, get help..

 

Sorry, 3 minutes and 30 seconds.

Come on. you posted a bunch a numbers, temps and values but they are worthless if you hide some values like time, power-consumption, and so on. Example from your log:



So what exactly says your log now? Your GPU temp reaches after 3 minutes an avg temp of 52°C, ok and now? Is that good or bad? Your GPU goes up to 100w, but the most of time it's only 54,48w. Your gpu temps are possibly there, because your gpu had only 50% load. Same for the clock. max 1950 Mhz, but avg only 1.050 Mhz. You can't judge your pch temps or other temps when your system runs with 50% load under 5 minutes. Nobody uses his notebook in a real-world test like gaming for 3 minutes or does everyday some benchmarks. Especially in gaming, where CPU and GPU are working a longer time. That's the purpose of these notebooks.

For a better comparison. In my FC5 tests i had with my weak GTX 1060 more load and higher clock and this over 30 minutes.



First do the same test or take some other game/ test with similar load and time and then you'll see the difference and a much higher pch temp.

No other arguments for the m15? This was my R5 btw. If you want to run your system stabil over 5.0Ghz and highest benchmark-scores you need these kind of mods but you don't have an HK-CPU, so you can't know that.

 

Your English is not great, and I'm actually not sure what your argument is anymore. What pertinent information are you missing exactly and why is it pertinent? You know what? Never mind.
I've seen enough evidence first hand that additional PCH cooling confers no immediate benefit in terms of basic benchmarks. A further discussion is unnecessary and thus I'd like to welcome you to my ignore list.

 

This guy really enjoys trash talking people/making up ridiculous stuff and then saying that other people are in fact the ones doing it, then blocks them lol. His English is great. It's not perfect, but can you also speak German and English very well? You can't even hold a debate /facepalm.

I appreciate your info and mod guides Sk0b0ld. One day I will build one of those expensive laptop coolers like you made but rn I don't need a 100$+ cooler lol

 

I wanna add, I have been testing the heatsink on PCH mod for about 2-3 days now and I no longer experience the random stuttering in games. I didn't record before and after results, which doesn't matter when there is only one single person in this thread that doesn't understand lower = better. But yeah, I used to stutter every 15min for about 5 seconds when playing CPU intensive games and it no longer happens. Whether specific people believe the PCH is throttling or not, I think it happens at 91C, that is my experience. The mod dropped it from 91C max to around 80-81 even though its covered in Kapton tape lol (wanted to use it asap and not wait a 24hr cure time for glue). I'll probably leave it this way until I need to open the laptop, since that is better than stock

 

I too have been experiencing random stuttering and freezing, even whilst in desktop. I am concerned over bigger issues, but I will get my MOBO replaced eventually.. still waiting on the parts to arrive in my country.

 

I am surprised even so many years later they still have a manufacturer for heatsinks. But yeah, I used to get them all the time in RB6 Siege and the new CoD. They use a lot of CPU. Even playing for hours straight (admittedly like 4-5) I haven't experienced any in since the PCH heatsink mod. Literally nothing else changed. The PCH turns red when it hits 91C or higher in HWinfo and that is when I noticed it would happen.

 

I don't know why, but my system actually encounters a BSOD when running HWinfo64 over a period of time. One of my sensors also turned grey during benchmarking after it hit 99 Degrees C. Well... At least the thermals are somewhat under control for now, after I undervolted both my CPU and GPU. I am hoping for better results once a DELL technician actually inspects the internals.

 

Maybe it has to do with some kind of USB device like a external HDD? Those are kept awake so long as the HWinfo is running, or at least mine are.

 

Adding this in the R1 thread as well to get opinions (recently received an R2, but has more extreme OLED vertical banding than I’d like so I’m considering a repair or just returning it for the R1).

********

I’ve recently put in a second order for an M15 R1 since it is considerably cheaper than the R2. Obviously pros and cons exist, but to get an OLED equipped M1 with nearly all the same specs I am saving 475 dollars (675 if I don’t count the VISA prepaid card as I haven’t received that yet).

SO R1 Specs in order just placed: 9750, WIFI6, 16GB, 2060, OLED, 90Wh, 512GB drive
My R2 currently in hand: 9750, WIFI6, 16GB, 2070 MQ, OLED, smaller battery, 2TB drive

What I gain going to the R1:

• Money in the pocket
• Better maintenance, and easy repaste
• Num-pad? Lol
• I think I prefer the R2 design language overall, but I don’t like the front edge of the R2 at all compared to the R1 (the glossy angled edge of the R2)
• I don’t have to worry about that beautiful white surface getting discolored on the R2
• 90Wh vs 76 seems like another hour

What I would lose going to the R1

• 2070 Max-Q performance - but I think this is only a 5-10% frames loss?
• Storage - although I plan to put my existing 2TB Intel SSD from my desktop in there tooffset, but I was going to put it in the R2 anyways so I don’t consider the price of that drive part of the delta
• Smaller bezels
• Less color calibrated OLED? I’m not a “creative” so I don’t think this matters

Inviting everyone into my brain - it’s tough but the banding on my R2’s OLED is probably one of the worse I’ve seen so I would have to either return it or try to have them swap out the display, but a tech mucking around and losing screws on my 2300 dollar laptop sounds no fun.

 

As a minor tweak, I'd recommend ordering the R1 with 8GB, and getting HyperX Impact RAM. Better RAM and money left in the pocket.

The 2070 MQ prob has the best perf/price. For battery use there is no difference, obviously, and for desk use you can upgrade to AGA to max out the GPU oomph.

 

Around the holidays you can buy a i1 Display Pro for about 100-130$ if you care a lot about a calibrated display. That is something you could even make some money off of or just make your investment back, though it only works on computers. I thought the m15 r1 1080p display was good out of the box, but it looked even nicer after being calibrated.

From what I remember, nearly everything is soldered on the r2's. Even the WiFi/BT chips lol. I think the only thing that isn't soldered, is maybe some NVME slots? That alone is extremely dumb if you don't have extended warranty, since it would be very hard to fix yourself if needed.

Small bezels are cool, but if you look at the dimensions it doesn't really reduce the size much. The bezels on the r1 are only a few mm (3-4?).

The r2 does look nicer for sure. Going from a 15R3, the Alienware logo lights up and you can choose the color. On this one it just looks so boring and plain, like it looks like its broken because its a plain silver logo. However, I prefer laptops to be less flashy... it makes people ask less questions.

If you ask me, I'd say go for the m15r1.
- Wait until January sales if you can, they are the best. If not, Christmas sales are better than black friday.
- Use cashback websites on Dell
- I recommend to buy used/open box/refurbished.
- Why should you buy refurbished? Because you save 300-500$ for it having a scratch or being used. You put that money towards a extended warranty (which also goes on sale around holidays) and now you have a 'used' laptop that if something happens, you get very good support and fre
- If you can find a website that sells Dell gift cards, you can save easily 8-10% off 2k this way too, but I think those are harder to find nowadays.
- If you are buying new, definitely make a post on some Dell Facebook/twitter pages asking for a dell brand ambassador. If there is no good cashback offerings, you can use a Dell brand ambassador to help both them in yourself, because you save 10%.
- make sure you sign up for one of those Dell account things so you get 3% cashback in the form of a dell gift card. I got a 90$ gift card, bought a Xbox one and then sold it.

If you do all of this, you can get a Dell laptop for about 30-50% off. Which is funny because everyone says Dell is so expensive... meanwhile you get a cheap laptop and the best + longest accidental damage support

 

Wow thanks for some of these inside tips. I’m active duty, so was taking advantage of the 15% discount this weekend. The R1 model price for the specs I describe is $1665, vs $2340 for the R2 I already received. That R2 also includes a $200 VISA prepaid card rebate thing (Haven’t submitted for it yet), but that’s future money. I also have DPA account and wanted to use the 12 month no interest payment thing as well, which I can’t use for outlet sales. I see outlet can save another 100-150, but I preferred new because DPA allows me to do this and return without seeing any charges on credit cards if need be. To be honest, I’ve usually used ebates cash back in the past on numerous purchases (6-10% usually), and after returning laptops, never really got back to ebates to say hey, I returned this. So I kind of owe Dell at this point, lol.

 

Also I should say, I agree with you on Dell costs - even a moderately savvy buyer can get better deals than MSRP. The reason I’m picking the “uglier” M15 over the Blade is simply because of Price. I would prefer the Razer Blade aesthetic, but no way I can get a 4K OLED Razer gaming laptop for 1650 bucks. Even if I compromise on my desire for 4K OLED, the Blade 15 with a 2060 is still 1900 dollars without any extended warranty. Both prices above I included a year of premium support.

 

Cashback websites offer as much as 12% cash back around holidays, usually on new systems. Maybe the prices changed more recently, but I feel like around the holidays the used/refurbished systems are even more discounted than they would normally be now.

I bought my first Alienware around Black Friday and ended up returning it after I saw all the sales January and bought a higher spec'd laptop for same price. But then I learned if you go with accidental damage warranty, you get a used product in return. So imo there is no point in paying for accidental damage + a new laptop because you won't get a new system in return if yours is damaged badly.

I always thought the veterans discount was a all-year-round thing from Dell and I don't think it would stack with other coupons. There really is a lot of variables with choosing what to go with. Because you can use a coupon + 12%cashback (I recommend just checking cashbackmonitor website) and that works. However You can't use a coupon and than a brand ambassador discount and I also don't think a veteran discount works with coupons.

Afaik, you can use 3 cards max to make a Dell purchase. So if that is 3x 1000$ dell gift cards, or 2x 1000 gift cards and credit, provided they didn't change it. The website Raise used to sell Dell cards often but they seemed to not have them lately.

Also to anyone that is curious because I was surprised, but when you go to extend a dell warranty it is actually only about 10% higher cost than when originally bought. It's about 1009.40$ for 5 more years, I guess provided you buy it before the warranty ends. Which is pretty crazy because if you would buy a laptop right now, it costs about 910$ for 5 years of accidental damage. It makes me like Dell even more.

Edit: apparently using warranty service resets the point of sale for the laptop and that is why the warranty prices I was given were so good. Didn't know until I asked the Dell guy to explain.

 

Good to know on extending a warranty - since I only bought the initial year, nice to know I can extend near the end of the term. I’m seeing that Dell is starting to offer monthly payment for warranties on the G5/G7 line, and XPS 15. Right now the monthly costs seem reasonable. EVERYTHING WILL BE SUBS in about 5 years. We won’t own anything anymore.

 

Does anyone know if the laptop CPU have any noticeable benefit from using 2400Mhz RAM vs 2666 Mhz? I stupidly bought a 1x 16GBHyperX 2400Mhz due to it having low timings 3 years ago since I wasn't aware higher speeds and timings worked better for gaming. Now I am wondering if I should buy another 16Gb Stick of 2666Mhz and just let it run @ 2400Mhz. Funnily most reviews compare 2133Mhz to 2666Mhz. Or 2400Mhz to 3000Mhz, they just kind of skip over one of the meaningful speeds. Either because its very insignificant or because the graphs look nicer with the bars having bigger gaps.

 

Silly question but does anyone have an eBay/Amazon link to where I can purchase both CPU and GPU fans for the alienware m15? Or even a part number? (Both fans are different from each other)

Sent from my Moto Z (2) using Tapatalk

 

Received an R1 yesterday - comparing to my R2 in hand. I’m surprised that now that I have the R1 in hand I prefer the more traditional design over the R2 in almost every way! I thought I would like the R2 more as it’s unique, but no, R1 is a better design. Keyboard, and per key RGB is nicer on the R2, along with trackpad. My R1 seems to have a somewhat loose feeling trackpad, but a repair on that looks really simple should I choose that.

The main thing I’m excited about is the R1’s OLED display does not exhibit serious vertical banding. See attached pictures. This is obviously a defective panel on the R2. I’ve been debating whether to live with it, but too many screens in windows use the exacerbating light gray color scheme (windows explorer, any dark theme in word, etc). I would share a video to show how distracting the vertical banding is on the R2 but I think file-size limits really prevent that.

My question to those in the know, the R1’s RED color trends towards orange, how do I correct this? I don’t own any color calibration tool.

 

Another question - my R2’s thermals appear to be better out of box. My GPU on the R2 (2070 MQ) runs at mid 70s under load for long periods of time. The 2060 in the R1 seem to be hovering in the mid-80s. Is this normal for the 2060, or is this some bad heat sink action going on? CPU between both is similar, although both will be in the mid 90s after undervolt and throttle quite often under AIDA64/Valley stress test. R2 CPU temps tend to be a few degrees cooler, but it’s so hard to compare as the system makes power limit decisions and the frequency adjusts all over the map. R1 was able to stick around 3.5 GHz, and 95C through a 20 minute AIDA/Valley stress test for the most part. Yes hot, but at least it seemed to maintain that frequency fairly solid.

Another quirk, R1 fans in “Balanced” mode in AWCC frequently adjust up and down in RPMs, the R2 was more stable - is this also normal fan behavior?

 

What is the max frequency clock you get for that 2070 Max q?

Sent from my Moto Z (2) using Tapatalk

 

I’d have to run the stress test again to see where the GPU parks itself at, I was noticing some ups and downs, but Valley stresses the GPU in a similar way since it changes “scenes” so often. Don’t have the R2 laptop at work right now.

R1 is running around 1545 MHz on the GPU though after 2+ hours of valley alone (no AIDA). Temp is 86C after that period.

Update: Adding in that lifting the back end up about .5-.75 inches and now GPU is at 78C after about 10 minutes and closer to 1600 MHz stable, CPU is cooler as well, <90. This is just a valley test, not joint AIDA/Valley. I know this has been reported elsewhere to really help thermals but wow, that’s a lot. I don’t really want to use a “laptop pad”, but does anyone know of a good pair of adhesive feet that can accomplish this same thing (or something that folds down) without necessitating bringing anything else around with me?

 

Re temps: delay a repaste with Phobya Nanogrease Extreme or LM no further. Tip: take extra care with the 7 heatink screws, they used a cheap alloy and they have propensity towards stripping.

 

I really enjoy this laptop stand - https://www.amazon.com/gp/product/B007BD16KC/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1 . Its silent and folds. It definitely helps temps too. I feel like its better than most stands, many laptops nowadays do a great job of circulating airflow and adding crappy fans onto a stand likely holds back the cooling or just adds noise without extra cooling. If I would get any laptop stand with fans I would make Sk0B0lds custom stand, since that has high quality fans in customization locations, its effective.

 

I ordered an M15 (R1) today.

Trying to get through all the info in this thread. There's a lot

I'm looking to replace the Toshiba drive with a 1TB 970 Evo Plus. Will install Windows Pro in it and just download the drivers from the Dell website.

Mines come with 32GB of Ram so no need to change the memory.

Any other worthwhile upgrades?

* Is it worthwhile upgrading the WiFi card?
* Is it better to remove the killer network software that is installed?
* Mine unfortunately comes with the 4 cell 60Whr battery. Not a huge problem as it will be mostly plugged in, but I'm curious as to whether the design is different in mine. Can the larger battery be used? Does the extra space allow a 2.5" sata SSD to be added?
* Is a repaste worth it? I saw someone saying that screw stripping is a risk. I've just sent a laptop back to dell as I couldn't take out the keyboard to replace the memory as the screws were on so tight they couldn't be moved a mm. So concious of the fact Dell uses poor screws.

Any other advice for a new/future M15 owner?

Sent from my ELE-L09 using Tapatalk

 

Repaste is absolutely worth it, but again need to be extra careful and use the correct screwdriver with those screws. Zero problem unscrewing but I would not try to super-overtighten these.

 

Oh man I'm scared after the precision laptop they sent me. I've got a tech screwdriver kit I've used for years and a few other dedicated M2 screwdrivers and nothing could budge them. I've used them with several motherboards and laptops over the last few years, and nothing worked with the Dell.

So are people seeing drops in temperature with the repaste of, say 5 or 10 degrees? Any particular paste everyone is using?


Just checked and I have Artic MX-2, but not sure it it's the best solution for for a mobile GPU.

Sent from my ELE-L09 using Tapatalk

 

I can only recommend one non-LM paste: Phobya Nanogrease Extreme. Very easy to apply and better results than Kryonaut.

 

Thanks for the suggestion. Only £10 too.

Had a look at an instruction video for it and it doesn't look too difficult. Once you get the heatsink out, it's the same process as a desktop repaste.

Appreciate your suggestion.

Worth changing the wireless card too?

Sent from my ELE-L09 using Tapatalk

 

Yes, the R1 is trivial to repaste. Not feeling a need to replace my Killer WiFi - works fine in Windows. Good luck and enjoy.

 

Thanks. Appreciate your suggestions.

Sent from my ELE-L09 using Tapatalk

 

Yep, i second that, i'm using exactly that one on my R1. Since i travel a lot, is extremely convenient and works really well.

 

Saw that one previously, although I’m not sure it meets my needs overall. I don’t really want to add more to the bag if I can avoid it. Although it is kinda easy to find any object that can prop up when needed. It wasn’t 100% necessary to do this on the R1, just seems to improve cooling and probably adds a bit of a performance boost overall. Not sure it will be make or break whether this thing runs at 78C/90C vs 86C/95C. I know some folks will come along and say it matter for “longevity” but in my research this isn’t substantiated.

 

I’ll add that the only thing that concerns me is any “lurching” in performance if the system is very poor at how it manages the power levels and thermals. I.e., if it causes lurches or massive frame time spikes in games. I’ve not tested my R1 enough to see if it is doing that poorly. Further, I think it’s easy to avoid it altogether if you are willing to accept a 36X max multiplier for 4+ cores and help keep the system from reaching it’s thermal/power limits.

I’m more than fine if this R1 sustains mid 3 GHz Frequency at 95-97C without major adjustments over hours of gaming. Obviously if it could get closer to 4 GHz that would be great, but I’m interested in consistency over time, not really that the CPU is staying below 90C. I’m also interested in quieter fans. The other manufacturers all have the same issues.

 

Just wanted to mention that with repaste, proper undervolting (including the static UV trick), and probably somewhat redundant RAM and PCH cooling, the machine is quite cool even in balanced (I posted benhhmsrk temps a few pages back). This has been tested in (too) long competitive CSGO sessions where any FPS drop under 200 results in instant death lol. No issues.

 

I’m pretty solid technically, but even I got dizzy looking at all the TS tweaks a few of you were describing way back on page 314 of this thread. Holy moly - is there kind of a “safe” guide to specific TS settings recommended? I really like the idea of just two profiles, one for mains, other for battery. Battery is easy to just turn off Turbo, but on the mains I just want one running. Right now I’m at -140mV undervolt and it works fine combined with a 36X max multiplier, but temps still over 90 during a stress test. I saw a lot of different settings discussed, so I wasn’t sure which one was the best to get the most frames out of this machine.

I also haven’t done any re-paste yet. Most effective thing I’m seeing so far is just lifting the back end off the table by 3/4 inch.

 

Elevating the back of the laptop is a must for all those ultrathin laptops if you are playing AAA games. I don't understand why they don't add foldable "legs" like the one in keyboards. It would be trivial and help a ton with cooling and ergonomics.

If you want to travel light I recommend you to get something like https://www.amazon.com/Ergonomic-Invisible-Portable-Magnetic-Computer/dp/B07NKP4MPZ/ They are very portable, and you can find them in AliExpress for about $5 if you are willing to wait.

 

Have a 90 min AIDA/Valley stress test going.

After 52 Mins/stock: CPU at just under 3.5 GHz @ 97-98C - stable
Switched to -125mV undervolt: CPU still at 3.5 GHz after 20 minutes @ 97-98C stable
Lifting the back end up 1”: CPU drops to 92C, 3.5GHz

For GPU, it’s 86C until the 1” lift, after which it drops to 82C - basically around 1545Mhz average in all cases (and fluctuates up and down).

I’ll try -140 mV, but I’m not getting this thing to care about my undervolt so far, and I’m not an idiot with TS. The same undervolt on my R2 saw a drop to around 90C but I haven’t stressed it as long as this. I’ll have to bring the R2 into work tomorrow to do the same thing to it. Might be a case where too much thermal mass is in the system since I spent 52 minutes at stock? Undervolt might simply be too little too late? Would be interesting to let it fully cool down and then start from scratch with undervolt applied.

Edit: -140 mV still not producing any change. It’s only lifting the back edge up that’s doing anything to get thermals down a bit. CPU performance seems unaffected in all cases, hovers at 34 watts regardless. GPU seems to like getting some more air in and looks more consistent when the back end is raised.